Efficient and Customizable Top-k Query on Named Data Networking

Offering a flexible paradigm for efficient data discovery is a crucial requirement in large information-centric networks. However, a major hindrance to query data in multicast networks is the uncertain sources and crowds of answering, which causes heavy network traffic and unduly long response time. In this paper, we present an efficient and customizable top-k query processing architecture, called NDN-Qk, by leveraging the emerging named data networking (NDN). NDN-Qk provides a completely ad-hoc and flexible query paradigm and enables the forwarders that distributed on NDN to cooperatively discover and aggregate the relevant top-k data. Specifically, we present adaptive types and strategies of data aggregation, and response completeness discerning algorithm for efficiently aggregating multiple top-k answers in NDN. Moreover, we further introduce a customizable query solution for adjusting the scale, score and range of response data in NDN-Qk. Finally, our comprehensive experimental evaluations demonstrates the effectiveness of our top-k query processing solutions and the performance of our proposed algorithms.